Application of Support Vector Machine (SVM) for Occupancy Status Data of Rehabilitation and Reconstruction Houses Post-Mount Merapi Eruption

Abstract

This research focuses on the application of the Support Vector Machine (SVM) classification algorithm to analyze the occupancy status data of rehabilitation and reconstruction houses (Rehab-Rekon) post-Merapi eruption. The eruption caused significant damage to residential areas, necessitating urgent rehabilitation and reconstruction efforts. The study aims to develop a predictive model that can classify the occupancy status of houses post-rehabilitation and reconstruction, aiding decision-making in aid distribution and reconstruction prioritization. The SVM algorithm was chosen for its ability to handle complex data and generalize well, thus improving the accuracy of occupancy status predictions. The dataset includes 2,516 houses, with 2,146 occupied and 370 with unclear occupancy status. The research seeks to validate the hypothesis that the SVM algorithm can effectively classify the occupancy status of Rehab-Rekon houses with a satisfactory accuracy of 91,27%. The findings are expected to demonstrate the potential of SVM in disaster recovery, particularly in the rehabilitation and reconstruction sector, providing valuable insights for stakeholders in planning and executing targeted and sustainable rehab-reckon programs.